Diagnostic tool: ICD-10

Inclusion criteria: - Aged 18-65 - Met ICD-10 criteria F20.0-20.3, 20.5-20.9 excluding acute psychosis - Duration of illness >=2 years - At least 3 items >=4 points in the PANSS Negative subscale.

Total sample size: No. randomised 153

Total sample size: ITT population 144

Gender: % female 38%

Age: Mean 40

Setting: Outpatient

Setting: Inpatient

Baseline stats: No significant differences between groups

Interventions Intervention - group 1.: Flupenthixol, 4-12mg/day, n=76

Intervention - group 2.: Risperidone, 2-6mg/day, n=77

Notes about the interventions: Dosage was adjusted as indicated.

Outcomes Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - PANSS, MADRS

Adverse events: Average score/change in specific adverse effects - ESRS

Satisfaction with treatment: Service user satisfaction - DAI

Quality of Life: Average score/change in quality of life - EuroQoL

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Not reported adequately

1.3 An adequate concealment method is used.: Not addressed

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Adequately addressed

1.6 The only difference between groups is the treatment under investigation.: Well covered

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Well covered

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was

completed?: 20-50%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). :

Adequately addressed

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Not reported adequately

2.1 How well was the study done to minimise bias?: +

Study ID KINON2006B

General info Funding source: Any pharmaceutical industry support

Published or unpublished data?: Published

Method Type of study: Individual randomised trial

Type of analysis: ITT - All participants with at least one post-baseline assessment

Type of analysis: LOCF

Blindness: Double-blind

Duration: No. weeks of treatment - 24

Raters: Not stated to be independent of treatment

Design: Multi-centre - US

Number of people screened, excluded & reasons: No mention

Notes about study methods: Patients were assigned to treatment groups based on a computer-generated randomization code. The randomisation was balanced by using permuted blocks and was stratified by site. All study medication was identical in appearance and was dispensed to subjects by study site personnel.

Participants Diagnosis: Schizophrenia [% of sample] 67%

Diagnosis: Other schizophrenia related [%] Schizoaffective bipolar 23% Schizoaffective depression 10%

Diagnostic tool: DSM-IV

Inclusion criteria: - Met DSM-IV criteria for schizophrenia or schizoaffective disorder - Met criteria for prominent negative symptoms, defined as a PANSS >=4 (moderate) on at least 3, or >=5 (moderately severe) on at least 2 of the 7 negative scale items; and for social and functional impairment, defined as a GAF <=60 (moderate difficulties).

Total sample size: No. randomised 346

Total sample size: ITT population - Varied

Total sample size: Safety population 346

Gender: % female 35%

Age: Mean 41

Ethnicity: White 52% African descent 37% Hispanic 9% Other 2%

Setting: Outpatient

History: [Olanzapine / Quetiapine] Age of psychosis onset: 24.16 (8.73) / 22.59 (7.62) Years of illness: 17.57 (9.65) / 17.78 (9.39)

Baseline stats: [Olanzapine / Quetiapine] GAF: 43.24 (8.71) / 43.19 (9.73) SANS: 61.4 (17.4) / 60.4 (18.4) PANSS Total: 84.1 (12.8) / 85.2 (14.8) CGI-S: 4.2 (0.6) / 4.3 (0.7)

Interventions Intervention - group 1.: Olanzapine 10-20mg/day, n=171

Intervention - group 2.: Quetiapine 300-700mg/day, n=175

Notes about the interventions: Patients’ current antipsychotic medications were tapered off as their study medication was initiated. During study period 3, patients were titrated up to their clinically optimal dose of study drug (OLZ, 10–20 mg/d in 5mg increments; QUE, 300–700 mg/d in 100-mg increments). Dosage increases could occur at 7-day intervals after visit 4. Dosage decreases could occur at any time; however, the dose could not decrease below 10 mg/d for OLZ or 300 mg/d for QUE. Patients who required more than 2 dose decreases or dosages less than the minimum allowed were discontinued from the study. Dosing was flexible, and investigators were encouraged to use the highest doses necessary in both treatment groups. All study medication was administered twice daily. Each patient was treated at community mental health centres and assigned case managers, who developed a 6-month treatment plan for illness management and recovery in collaboration with the patient.

Outcomes Leaving the study early: Leaving because of adverse effects

Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Global state & service outcomes (e.g. CGI): Average score/change in global state - CGI

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - SANS, PANSS, CDS, CMRS+

General and psychosocial functioning (e.g. SFS): Average score/change in general functioning - GAF, PFQ, CMRS+

Adverse events: Average score/change in specific adverse effects SAS, BAS, AIMS, use of anticholinergic medication Laboratory tests, weight, BMI

Adverse events: Number of people with specific adverse effects Various

Adverse events: Number of people with general adverse effects

Quality of Life: Average score/change in quality of life QLS

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Well covered

1.3 An adequate concealment method is used.: Well covered

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Well covered

1.6 The only difference between groups is the treatment under investigation.: Adequately addressed

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Well covered

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was

completed?: >50%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). : Well covered

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Not reported adequately

2.1 How well was the study done to minimise bias?: ++

Study ID LINDENMAYER2007

General info Funding source: Any pharmaceutical industry support

Method Type of study: Individual randomised trial

Type of analysis: ITT - For all patients who did not complete the entire study, a likelihood-based repeated measures model (mixed models repeated measures) was applied.

Type of analysis: LOCF - applied to PANSS data only if patients completed >= 8 weeks of the study

Blindness: Double-blind

Duration: No. weeks of treatment 12

Raters: Not stated to be independent of treatment

Design: Single-centre - State psychiatric hospital, New York, US

Number of people screened, excluded & reasons: 36 participants were enrolled in the study, 35 were randomly assigned. One patient did not receive study treatment due to withdrawal of consent.

Notes about study methods: Randomisation procedure not reported

Participants Diagnosis: Schizophrenia [% of sample] 100%

Diagnostic tool: DSM-IV

Inclusion criteria: - aged 18-60 - PANSS total score >=50, with a PANSS negative subscale score >=20. The negative symptom score was required to contain >=3 items scores of >=3 - All participants fulfilled the criteria for the Schedule for the Deficit Syndrome (SDS) which included negative symptoms that are stable rather then unstable-state manifestations.

Exclusion criteria: - PANSS depression item score >=4, PANSS positive symptom subscale score of >=20 - SAS score >=2 - History of treatment failure on antipsychotics (persistent positive symptoms after 8 weeks of treatment with adequate doses of 1 or more antipsychotics.) - Significant medical disorder - positive substance misuse diagnosis within the last 3 months. - Pregnant or breastfeeding women and women of childbearing age not using adequate contraception

Total sample size: ITT population Unclear

Total sample size: No. randomised 35

Gender: % female 6%

Age: Mean 39

Ethnicity: White - 6% African American - 78% Hispanic - 13% Other - 3%

Setting: Inpatient

Setting: Outpatient

Baseline stats: [Haloperidol / Olanzapine]

PANSS total: 70.79(9.86) / 71.25(17.46) HAM-D: 5.58(3.13) / 5.74(4.00)

Notes about participants: [Haloperidol / Olanzapine] Prior antipsychotic treatment, n Haloperidol: 4 / 2 Thiothixene: 1 / 1 Olanzapine: 2 / 0 Risperidone: 8 / 10 Thioridazine: 1 / 0 Fluphenazine: 2 / 2 Aripiprazole: 1 / 0 Ziprasidone: 1 / 0 Quetiapine: 2 / 1 No previous antipsychotic: 0 / 1

Interventions Intervention - group 1.: Haloperidol, 15-20 mg/day; n=19

Intervention - group 2.: Olanzapine, 15-20mg/day; n=16

Notes about the interventions: Patients on antipsychotics decanoate preparations prior to the study were converted to oral tablets at equivalent doses at least 3 weeks prior to entry. - participants started with a fixed dose of olanzapine (15mg/day) or haloperidol (15mg/day) for the first 6 weeks after 1 week of cross-titration from previous medication. - fixed dose period was followed by a 6-week double-blind, flexible dose phase. - dose of medication could be increased or decreased by 5mg/day at 2-week intervals during the second phase to a maximum of 20mg/day in both groups. Study-dose change was based on a lack of improvement in PANSS negative symptoms. Haloperidol - Mean dose at end of study = 17.11(3.84) mg/day - participants received additional blinded active benztropine mesylate 2mg PO b.i.d Olanzapine - Mean dose at end of study = 18.44(2.39) mg/day - participants received benztropine mesylate placebo tablets.

For all participants, if significant EPS persisted despite benztropine, the dose of study drug was decreased. If this did not work, 2mg/day could be added in all cases.

Outcomes Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Mental state (e.g. BPRS, PANSS, BDI): Clinically significant response in mental state - defined as a 20% decrease of the PANSS negative subscale score

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - PANSS; HAM-D

Adverse events: Average score/change in specific adverse effects - SAS; AIMS

Cognitive functioning: Average score/change in cognitive functioning Test batteries were created for: Executive functioning, Declarative verbal learning memory, Attention and processing speed, Motor functioning

Other: Weight change, vital signs, laboratory values and prolactin levels

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Not reported adequately

1.3 An adequate concealment method is used.: Not addressed

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Adequately addressed

1.6 The only difference between groups is the treatment under investigation.: Adequately addressed

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Adequately addressed

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was completed?: <20%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). : Adequately addressed

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Not applicable

2.1 How well was the study done to minimise bias?: +

Study ID OLIE2006 [TA: Study 128-305]

General info Funding source: Not mentioned

Published or unpublished data?: Published

Method Type of study: Individual randomised trial

Type of analysis: ITT - All randomised participants who had a baseline and at least one post-baseline efficacy evaluation, regardless of whether protocol inclusion/exclusion criteria were met. - Evaluable population comprised randomised participants who had >=4 weeks of double-blind treatment and no major protocol deviations or violations.

Blindness: Double-blind

Duration: No. weeks of treatment 12

Duration: Median treatment duration - 12 weeks,

Raters: Not stated to be independent of treatment

Design: Multi-centre - 26 centres in Western Europe

Number of people screened, excluded & reasons: 143 participants screened, 20 excluded due to failure to fulfil inclusion criteria.

Notes about study methods: Randomisation procedure not reported

Participants Diagnosis: Schizophrenia [% of sample] 100%

Diagnostic tool: Other DSM DSM-IIIR

Inclusion criteria: - 18-64 years with a primary diagnosis of schizophrenia. - Indication for maintenance therapy with antipsychotic medication. - Women were either not of child-bearing potential or were practicing contraception. - Baseline scores on the Negative PANSS subscale had to exceed the PANSS positive subscale by =>6

Exclusion criteria: - Acute exacerbation of schizophrenia or psychosis 12 weeks prior to screening. - history of psychosurgery - severe medical illness

Total sample size: ITT population 122

Total sample size: No. randomised 123

Total sample size: Safety population 123

Gender: % female 36%

Age: Mean 39

Age: Range 21-65

Ethnicity: 100% white

Setting: Outpatient

History: [Ziprasidone / Amisulpride] Mean no. of previous hospitalisations: 4.3(7.5) / 7.3(9.8) Mean duration of most recent hospitalisation (months): 4.0(12.4) / 1.6(2.7) use of anticholinergic: 53% / 49%

Baseline stats: [Ziprasidone / Amisulpride] PANSS Negative subscale: 31.03 / 29.00

Notes about participants: - Participants underwent a minimum 3-day run-in period for screening procedures, including both psychiatric and medical evaluations. - Permitted concomitant medications were lorazepam and temazapam. Anticholinergics and propranolol were gradually withdrawn (25% dosage reduction per week) but were reinstated, if needed.

Interventions Intervention - group 1.: Ziprasidone (80-160 mg/day) - Participants were started on 20 mg b.i.d.; after 2 days, dosage was increased to 40 mg b.i.d. At the investigator's discretion, dosage could be increased to 60 mg b.i.d. from week 2 onwards or to 80 mg b.i.d. from week 3 onwards. - Mean dose 118mg/day - n=60 (ITT n=59)

Intervention - group 2.: Amisulpride (100-200 mg/day) - The starting dosage was 50 mg b.i.d. This could be increased to 150mg per day from week 2 onwards and to 100 mg b.i.d. from week 3 onwards, according to clinical response. - Mean dose: 144.7 mg/day - n=63

Notes about the interventions: Doses were administered in three capsules approximately 12 hours apart.

Outcomes Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Leaving the study early: Leaving because of adverse effects

Global state & service outcomes (e.g. CGI): Clinically significant response in global state: CGI - responder was defined as having an observed score of 1-2 on the CGI-I scale at the last observation.

Global state & service outcomes (e.g. CGI): Average score/change in global state - CGI; GAF

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - BPRS; PANSS Negative subscale

Mental state (e.g. BPRS, PANSS, BDI): Clinically significant response in mental state - PANSS - responder defined as having at least a 20% decrease in PANSS negative subscale score at the last observation relative to baseline.

Responder rates based on 30-50% decrease in PANSS Negative scores were also calculated.

Adverse events: Number of people with general adverse effects

Adverse events: Number of people with specific adverse effects

Adverse events: Average score/change in specific adverse effects - SAS; BAS; AIMS; MDBS

Other: - ECGs, BMI, and clinical laboratory tests including blood cell counts, blood biochemistry and urinalysis were also conducted. - Clinically significant changes in BMI (defined as a change of >=7% were also reported.

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Not reported adequately

1.3 An adequate concealment method is used.: Not addressed

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Adequately addressed

1.6 The only difference between groups is the treatment under investigation.: Not addressed

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Well covered

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was

completed?: 20-50%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). : Well covered

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Not addressed

2.1 How well was the study done to minimise bias?: +

Study ID RIEDEL2005

General info Funding source: Any pharmaceutical industry support

Published or unpublished data?: Published

Method Type of study: Individual randomised trial

Type of analysis: Completer

Type of analysis: ITT All patients randomised with baseline data and at least one post-baseline measure

Type of analysis: LOCF

Blindness: Double-blind

Duration: Mean duration (for each group) - Mean treatment duration was 66.1 days in the quetiapine group and 62.7 days in the risperidone group.

Duration: No. weeks of treatment 12

Raters: Not stated to be independent of treatment

Number of people screened, excluded & reasons: - Upon entering the study, a thorough medical and psychiatric history was carried out. - No participants were excluded after screening

Notes about study methods: Randomisation procedure not reported

Participants Diagnosis: Schizophrenia [% of sample] 100

Diagnostic tool: DSM-IV

Diagnostic tool: ICD-10

Inclusion criteria: - CGI score >=4 - Presenting with predominantly primary negative symptoms according to PANSS - PANSS negative subscale score >=21 and at least 1 point greater than their positive subscale score

Exclusion criteria: - drug or alcohol misuse/dependence - suicidal tendencies - laboratory or ECG/ EEG abnormalities (blood or urine values outside standard range by more than 20%) - pregnancy or lactation - significant medical history (brain surgery, unstable somatic conditions, HIV +) - treatment with clozapine within 4 weeks of enrolment.

Total sample size: ITT population - ITT population not reported. Number of completers for each intervention group reported: [Quetiapine / Risperidone] No. of completers: 13 / 12

Total sample size: No. randomised 44

Gender: % female 39%

Age: Mean [Quetiapine / Risperidone] Age: 30.6(10.9) / 39.3(12.3) *statistically significant difference between the two groups at baseline.

Setting: Outpatient

Setting: Inpatient

History: [Quetiapine / Risperidone] Age of onset: 25.3(10.5) / 36.9(17.7) Duration of illness: 5.4(7.5) / 2.5(12.7)

Baseline stats: [Quetiapine / Risperidone] PANSS: 103.4(16.4) / 97.8(16.9) SANS: 66.7(20.6) / 51.7(21.1) SAS: 0.2(1.1) / 0.5(1.3)

Interventions Intervention - group 1.: Quetiapine,50-600 mg/day, n=22

Intervention - group 2.: Risperidone, 2-6mg/day, n=22

Notes about the interventions: Each participant underwent a 2 day washout period before beginning the trial. Quetiapine: 50mg on day 1, 100 mg on day 2, and then daily 100 mg increments to 600 mg/day on day 7. Thereafter, the dose of study medication was adjusted according to the clinical judgement of the investigators, with the maximum dose allowed: 800mg/day - Mean dose: 589.7mg/day Risperidone: initiated at 2 mg/day on days 1 and 2, increasing to 4 mg/day on days 3–5 and 6 mg/day on days 6 and 7. Thereafter, the dose of study medication was adjusted according to the clinical judgement of the investigators with the maximum dose allowed: 8 mg/day -Mean dose: 4.9mg/day -Besides standard clinical management, no additional psychotherapy was performed. -Both risperidone and quetiapine were packed in lactose capsules containing 100 mg quetiapine or 1mg risperidone, with identical size and appearance to maintain blindness. -Lorazepam (≤ 4 mg/day); zopiclone (≤ 15 mg/day); Biperiden hydrochloride (≤ 8 mg/day) were all allowed throughout the trial.

Outcomes Leaving the study early: Leaving because of adverse effects

Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Global state & service outcomes (e.g. CGI): Average score/change in global state - CGI

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - PANSS; SANS

Adverse events: Number of people with general adverse effects

Adverse events: Average score/change in specific adverse effects - SAS

Adverse events: Number of people with specific adverse effects

Other: ECGs, assessment of vital signs, weight gain, serum prolactin levels, cortisol levels.

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Not reported adequately

1.3 An adequate concealment method is used.: Not addressed

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Well covered

1.6 The only difference between groups is the treatment under investigation.: Adequately addressed

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Well covered

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was completed?: 20-50%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). : Well covered

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Not addressed

2.1 How well was the study done to minimise bias?: +

Study ID RUHRMANN2007

General info Funding source: Any pharmaceutical industry support

Published or unpublished data?: Published

Method Type of study: Individual randomised trial

Type of study: Individual randomised trial (Noninferiority/equivalence)

Type of analysis: LOCF

Type of analysis: ITT - subjects with at least one post-randomisation observation. Valid for efficacy sample - as used for the primary outcome parameter and included subjects with a minimum treatment duration of 8 weeks.

Safety population - all patients with at least one study drug administration after randomisation.

Type of analysis: Completer

Blindness: Double-blind

Duration: No. weeks of treatment up to 25 weeks

Raters: Not stated to be independent of treatment

Design: Multi-centre - 27 centres in Germany, 3 in Austria

Number of people screened, excluded & reasons: Not reported

Notes about study methods: Randomisation procedure not reported.

Participants Diagnosis: Schizophrenia [% of sample] 100% (ICD-10 diagnosis F20.0–F20.3, F20.5–F20.9)

Diagnostic tool: ICD-10

Inclusion criteria: - aged 18 - diagnosis of schizophrenia according to ICD-10 criteria for >=2 years - >=3 PANSS negative syndrome subscale scoring ≥4 points; - stable clinical state, e.g. maintenance treatment had been started or that consideration of a change in stable medication was not due to any acute exacerbation of positive symptoms.

Exclusion criteria: - contraindication to treatment with, or hypersensitivity to any of the study drugs - dependence on alcohol or illegal drugs according to ICD-10 criteria; - concomitant treatment with lithium, carbamazepine other mood stabilisers or other psychopharmacological drugs - treatment with flupentixol or risperidone within 4 weeks preceding study - history of treatment with clozapine (to avoid inclusion of treatment-resistant cases); - concurrent clinically relevant physical conditions; - acute suicidal ideation; - participation in a clinical study - pregnant or breast-feeding females, or those of child bearing potential not using a medically approved contraceptive.

Total sample size: Safety population 153

Total sample size: ITT population 144

Total sample size: No. randomised 153

Gender: % female 37.5% (Details for the ITT population only)

Age: Mean 40.39(11.98)

Ethnicity: Not reported

Setting: Inpatient

Setting: Outpatient

History: [Flupentixol / Risperidone] Years since diagnosis: 11.28(9.98) / 11.50(10.07) Number of previous episodes: 4.87(4.48) / 4.73(4.38)

Baseline stats: [Flupentixol / Risperidone] PANSS neg: 27.67(5.44) / 27.65(5.40) PANSS pos: 17.72(4.47) / 14.65(5.45) PANSS gen: 40.90(10.977) / 39.47(9.93) The above scores are based on a 3-factor solution used in the paper.

Notes about participants: All participants belonged either to the ‘minus’ or to the ‘mixed’ subtype assuring a significant level of negative symptoms. Treatment with antipsychotics before inclusion in to the study. [Flupentixol / Risperidone] Butyrophenone per os: 32 / 24 Phenothiazine per os: 11 / 8 Haloperidol depot: 3 / 6 Zuclopentixol depot: 3 / 2 Olanzapine: 9 / 11 Amisulpride: 3 / 3 Sertindole: 0 / 2 Others: 8 / 9 No current pre-study medication: 6 / 6 Unknown: 1 / 6

Interventions Intervention - group 1.: Flupentixol: 4-12 mg/day, n=76

Intervention - group 2.: Risperidone, 2-6 mg/day, n = 77

Notes about the interventions: The first week was a run-in phase, previous medication was washed out and study medication was given at the minimal dosage. - Thereafter, dosing for both study drugs was flexible within the specified ranges. - Medication was administered in identical capsules containing either 2 mg of flupentixol or 1 mg of risperidone. Drugs were

given twice a day (day one 2×1, day 2 up to 2×2, from day 3 up to 2×3 capsules). - The only permissible concomitant medications were anticholinergic agents (biperiden) and short-term benzodiazepines and non- benzodiazepine hypnotics (for sleep induction).

Outcomes Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Leaving the study early: Leaving because of adverse effects

Global state & service outcomes (e.g. CGI): Relapse - Defined as a deterioration of >=10 points on BPRS

Global state & service outcomes (e.g. CGI): Average score/change in global state CGI

Mental state (e.g. BPRS, PANSS, BDI): Clinically significant response in mental state Defined by a standard criterion as a reduction of baseline scores >=20% and by a strong criterion as a reduction of >=50%

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state PANSS - conducted a factor analysis, with resulting 3 and 5-factor models, MADRS

Adverse events: Number of people with general adverse effects

Adverse events: Average score/change in specific adverse effects ESRS

Adverse events: Number of people with specific adverse effects Table reporting all AEs experienced by >5% of the study population.

Other: BMI, vital signs including diastolic blood pressure and heart rate

Quality 1.1 The study addresses an appropriate and clearly focused question.: Well covered

1.2 The assignment of subjects to treatment groups is randomised.: Not reported adequately

1.3 An adequate concealment method is used.: Not addressed

1.4 Subjects and investigators are kept ‘blind’ about treatment allocation.: Well covered

1.5 The treatment and control groups are similar at the start of the trial.: Well covered

1.6 The only difference between groups is the treatment under investigation.: Adequately addressed

1.7 All relevant outcomes are measured in a standard, valid and reliable way.: Adequately addressed

1.8 What percentage of the individuals or clusters recruited into each treatment arm of the study dropped out before the study was completed?: 20-50%

1.9 All the subjects are analysed in the groups to which they were randomly allocated (often referred to as intention-to-treat analysis). : Well covered

1.10 Where the study is carried out at more than one site, results are comparable for all sites.: Adequately addressed

2.1 How well was the study done to minimise bias?: +

Study ID SIROTA2006

General info Funding source: Any pharmaceutical industry support

Published or unpublished data?: Published

Method Type of study: Individual randomised trial

Type of analysis: LOCF

Type of analysis: ITT all those patients who were randomised with baseline data and had at least one post-baseline measurement

Blindness: Only raters blind

Duration: No. weeks of treatment 12

Raters: Independent of treatment

Design: Single-centre Israel

Number of people screened, excluded & reasons: screening method not reported

Notes about study methods: randomisation procedure not reported

Participants Diagnosis: Schizophrenia [% of sample] 100%

Diagnostic tool: DSM-IV

Inclusion criteria: - primary enduring negative symptoms according to Carpenter's Criteria for the Deficit Syndrome. -negative symptoms not adequately responded to previous medication, defined as a lack of response to at least two conventional antipsychotics at a dose of 400-600mg chlorpromazine equivalents for a period of 4-6 weeks - PANSS negative subscale score >15 - SANS total >60

Exclusion criteria: - concurrent Axis 1 SAM-IV disorder - history of seizure disorder, or any clinically significant medical condition that would interfere with evaluations of efficacy or tolerability. - pregnancy - use of depot antipsychotic within one dosing interval - participation in another investigational drug trial within 30 days on enrolment

Total sample size: No. randomised 40

Total sample size: ITT population 40

Gender: % female 20%

Age: Mean 37

Ethnicity: Not reported

Setting: Inpatient

History: [Quetiapine / Olanzapine] Mean duration of illness, years: 15.9(9.1) / 13.3(7.4)

Baseline stats: Baseline not reported (Change from baseline used in the analysis)

Notes about participants: [Quetiapine / Olanzapine] Previous antipsychotic, n(%) haloperidol: 5(26.3) / 5(23.8) perphenazine: 4(21.0) / 4(19.0) sulpiride: 1(5.2) / 2(9.5) zuclopenthixol: 3(15.7) / 3(14.2) risperidone: 6 (31.5) / 7(33.3)

Interventions Intervention - group 1.: Quetiapine, 200-800 mg/day, n=19

Intervention - group 2.: olanzapine, 5-20mg/day, n=21

Notes about the interventions: Quetiapine - 50mg/d on day 1, 100mg/d on day 2, 200mg/d on days 3-4 and 300mg/d on days 5-7. Thereafter patients received 400mg/d for 2 weeks followed by 600mg/d for 6 weeks. - For patients who did not respond sufficiently to 600mg/d, dose increased to 800mg/d until the end of study. Olanzapine - 5mg/d on days 1-5, 10mg/d on days 6-10. Thereafter 15mg/d for 4 weeks. - For patients who did not respond sufficiently to 15 mg/d, dose increased to 20mg/d until the end of the study Patients in both groups were flexibly dosed throughout the study according to clinical response. Although patients were not receiving any other antipsychotic, biperiden was allowed for the management of EPS

Outcomes Leaving the study early: Leaving due to any reason (non-adherence to study protocol)

Mental state (e.g. BPRS, PANSS, BDI): Average score/change in mental state - PANSS; SANS

Adverse events: Number of people with general adverse effects

Adverse events: Number of people with specific adverse effects - Lists AEs occurring in >=7% of patients,

Adverse events: Average score/change in specific adverse effects - SAS; BAS; AIMS